Pesticide Environmental Fate - American Chemical Society

Chapter 4

Downloaded by UNIV OF PITTSBURGH on December 6, 2015 | http://pubs.acs.org Publication Date: March 1, 2002 | doi: 10.1021/bk-2002-0813.ch004

An Integrated Approach for Quantifying Pesticide Dissipation under Diverse Conditions I: Field Study Design 1

2

3

S. A. Cryer , P. N. Coody , and J. White 1

Dow AgroSciences, 9330 Zionsville Road, Indianapolis,IN46268 Bayer Corporation, 8400 Hawthorn Road, Kansas City,MO64120 Stone Environmental Inc., 58 East State Street, Montpelier,VT05602 2

3

Environmental fate studies for pesticide registration have traditionally focused on a single environmental media and over a single growing season (e.g., pesticide dissipation in the upper 90 centimeters of soil, residue levels in specific receiving waters, or residues leaching to ground water). Studies focusing on a single element of the hydrologic process, and under a narrow range of climatological conditions, can often generate more questions than answers in understanding and describing the environmental fate of a pesticide. These studies may not address the fundamental issue of mass balance closure for the pesticide and thus do not provide the necessary framework for predicting pesticide behavior under different environmental or climatological conditions. A n alternative study documenting chlorpyrifos dissipation in soil, foliage, runoff, and receiving waters under a range of agronomic and climatological conditions is presented. This multi-year integrated field study consists of a 17.29 acre corn production watershed near Oskaloosa, IA which drains into a 0.6 acre pond. In addition, numerous

© 2002 American Chemical Society In Pesticide Environmental Fate; Phelps, W., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.


44

0.16 acre meso-plot experiments are also subject to natural and/or simulated-rainfall and agricultural practices under a carefully controlled environment. Integrated studies quantifying runoff, dissipation, and scaling factors offer the advantage of a comprehensive, time- and cost-effective approach for evaluating the environmental fate of pesticides under field conditions. This rich data set is used for model validation and subsequent extrapolation to other environmental and climatological conditions to expand our understanding of the environmental impact resulting from the use of chlorpyrifos.

Introduction This four part study details the results of an integrated field study design and modeling program used to monitor chlorpyrifos dissipation in soil, foliage, runoff, and receiving waters under a range of agronomic and climatological conditions. The integrated design incorporates a series of meso-plots (0.16 acres), in which simulated rainfall and agricultural practices under a carefully controlled environment can be varied to investigate pesticide transport over a short time frame (i.e., single precipitation event). The integrated study design offers the advantages of a comprehensive, time- and cost-effective approach to evaluate the environmental fate of pesticides under field conditions. This is contrasted with the current U S E P A FIFRA Subdivision N study requirements that include numerous distinct and unrelated media-specific studies. The integrated study was designed to generate data required to calibrate/validate simulation models in addition to providing site-specific behavior for chlorpyrifos formulations. Numerical models can subsequently be used to evaluate other environmental and climatological conditions to provide a decision support mechanism for environmentally-sound product use. This document discusses the field study design. Details of field observations, model validation, and extrapolation procedures are discussed in the companion documentation (1-3).

In Pesticide Environmental Fate; Phelps, W., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

45

Methods And Materials A 17.29 acre watershed located near Oskaloosa, Iowa was treated with chlorpyrifos during the 1992-1993 corn growing season and monitored for offsite movement of the pesticide. Much of the study design followed traditional practices as outlined by Wauchope et al. (4). The watershed consists predominantly of silt loam (Hydrologic Group C) as seen in Table I.


Table I. Watershed Soil Properties Soil Series Ladoga silt loam Downs silt loam Fayette silt loam

Map Unit 76C2 T162B 163C2

Hydrologic Group C B B

Approximate Area (%) 70 15 15

The treatment of the watershed was typical of corn agronomic practices for the Midwestern U . S . A with corn planted on the contour. This particular watershed drained into a 0.60 acre farm pond. Both the watershed and pond (Figure 1) were surveyed and pond bathometric measurements made. Representative slopes within the watershed range between 2.6 - 5.1%, and the two primary drainage channels seen in Figure 1 have slopes of 2.6 and 3.2%, respectively. The pond had a single tile drain inflow (Station 6) and a primary drain outflow (Station 4), both of which were monitored for water flow and chlorpyrifos residues for mass balance closure. Time-dependent samples of vegetation, soil, pond water and pond sediment were analyzed for chlorpyrifos residues to characterize chlorpyrifos leaching and dissipation patterns. In addition, spatial scaling issues between meso-plots (ca. 0.16-acre) and the watershed (17.29-acre) were addressed by two nested meso-plots within the watershed. These nested meso-plots received the same management practices and natural precipitation as the watershed (Figure 1) with the only major difference being in surface area and length scale. In addition, four separate experiments consisting of 0.16-acre "meso-plots" were located within the same commercial cornfield as the main watershed, but part of a different drainage basin. These four artificially irrigated meso-plots each received different chlorpyrifos treatments at different times during the growing season to investigate both bare soil runoff and the effect of crop cover on hydrology, erosion, and chlorpyrifos edge-of-field transport.



46

Figure J. One foot topographic contour survey of watershed with locations for sampling equipment and nested meso-plots.

Instrumentation of the Main Watershed

Runoff Monitoring Stations Primary runoff stations (Stations 2-3) were located immediately northwest and southwest of the pond and were placed where the two primary drainage


47


channels of the watershed enter the pond. These stations were equipped to monitor runoff flow rates through the flumes (Figure 2).

Figure 2. Location of runoff monitoring stations placed around the pond perimeter. Sampling electronics in the primary runoff stations collected both a composited, flow-proportional sample of runoff where approximately 1 L of 3

runoff water was collected for each 750 ft of runoff passing through the Parshall flumes. Discrete runoff samples were also obtained according to a predetermined time sequence where the sampling frequency decreases as the runoff event continues. The discrete runoff-sampling scheme provided a capacity to sample long runoff events using the 24 sample containers supported by Isco model 2700 pump samplers. Flow-proportional samples were composited into stainless steel, 55-gallon drums while discrete samples were collected into 1-quart (946 mL) glass or metal containers (Figure 3). The sampling scheme allowed for intensive sampling of short-duration runoff events or during the initial stages of longer storms until sample containers were replaced. A l l sampling equipment was housed in a temporary shed to avoid possible malfunctions due to inclimate weather.



48

Figure 3. Primary flume sampling equipment comprised oflsco sampling pumps for both discrete and flow proportional composite sampling and a Campbell 21X data logger. A l l watershed runoff with eroded sediment was forced through the sampling flumes. Concrete footings and block were used for each primary sampling flume to avoid possible washout from extremely large runoff events (Figure 4). The primary flumes were sized using predictions from the U S D A model E P I C W Q (5) for l-in-10 year storm intensities for Southern Iowa. A single secondary runoff station equipped to collect composite, flowproportional samples from a relatively small drainage area within the watershed (identified as station 1) was located near the south shore of the pond (Figure 5). A Campbell Scientific 21X data logger was used at all sampling locations to control the sampling equipment and log water flow rates. The water level in each flume was monitored continuously using a Druck P D C R 950 submersible pressure transducer positioned in a stilling well connected to the flume.

Pond Overflow Monitoring The pond was constructed with an overflow drainpipe. Flow-proportional samples from this overflow drain were collected into a stainless steel, 55-gallon


49


drum using equipment identical to that of the secondary flume station. Following the sample collection, the compositing barrel was emptied and prepared to receive additional samples.

Figure 4, Concrete block retaining walls and primaryflumesampling stations for quantifying runoff hydrology, sediment yield, and chlorpyrifos transport. Pond Water and Sediment Sampling Pond water and sediment were sampled on both a predefined basis and defined time weighted intervals following any runoff event. Sample intervals were shorter immediately following a runoff event to properly capture chlorpyrifos dissipation patterns in pond matrices. The pond was sectioned into three zones for sampling purposes and samples from each zone were withdrawn on each sampling date (Figure 6). Water samples were obtained using a Sub-Surface Grab Sampler (Forestry Suppliers, Inc., Jackson, MS) with an aliquot collected just below the water surface, approximately half way to the pond bottom and near the pond bottom. The three aliquots were composited into glass or metal containers and were considered representative of a water sample from the entire water column for each specific section sampled.



50

Figure 5. Secondary flume sampling station. Nested meso-plots have similar sampling and electronic equipment (with smaller trapezoidal flumes).

Pond sediment samples were obtained using a Wildco 2410 sediment sampler with a 48 mm diameter cutting tip (Forestry Suppliers, Inc., Jackson, MS), which collects sediment cores into plastic sleeves. The sampler was forced into the pond floor to the maximum depth possible (~ 12 inches) by applying downward pressure. The core was then recovered from the sampler assembly and the plastic sleeve containing the sample was capped and maintained in a vertical position until frozen. Four sediment cores were generally collected from random locations within each of the three sampling zones to produce a total of 12 cores on each sampling day.

Instrumentation of Nested and Artificially Irrigated Meso-plots

Nested Meso-plots Within Watershed Meso-plots of dimensions 41 ft x 170 ft (- 0.16 acres) were nested within the main watershed and were instrumented immediately after the first



51

chlorpyrifos application was made. This allowed for the use of a commercial planter (John Deere 7200) without running the risk of interference by the scientific equipment necessary to quantitate runoff. A solid retaining wall was installed against the undisturbed soil with the upper edge approximately even with the soil surface. A metal gutter was then mounted to the retaining wall on a slight grade oriented perpendicular to the fall line of the plot. The gutter assembly attached directly to the inlet face of a large, 60° trapezoidal flume that was used to measure the flow from each meso-plot (Figure 7). Nested mesoplots were subjected to the same natural precipitation patterns as the watershed. Sprinkler irrigation was only added to those meso-plot experiments performed outside of the watershed.

Figure 6. Schematic representation for pond and watershed sampling zones for pond water, pond sediment and surface soil transects.

Flow from the nested meso-plots was continuously monitored using calibrated Isco model 3230 bubble flow meters in conjunction with a 60° trapezoidal flume. Runoff water was sampled from behind the flume using Isco model 2700 or 3700 pump samplers. The flow-proportionally-sampled runoff 3 (1-liter sample for each 3 ft of runoff) was withdrawn through a Teflon-lined sampling tube (3/8 i.d.) and was delivered into stainless steel, 55-gallon H


52


drums. Multiple aliquots for chlorpyrifos residue and sediment yield determination were withdrawn from the drum at the conclusion of each runoff event.

Figure 7. Schematic of Meso-plot and Drainage Area.

Meso-plots Outside Watershed Subjected to Artificial Precipitation Four artificially irrigated meso-plots were located outside the watershed in 1992 and were identical in terms of size, construction, and sampling electronics to the nested meso-plots listed above with the exception of irrigation. Irrigation sprinkler heads were mounted on vertical risers to mimic typical raindrop energies upon impact with soil (Figure 7). Irrigated meso-plots 1 and 2 had discrete samplers, while irrigated meso-plots 3 and 4 consisted of both discrete and flow proportional composite sampling. Pre- and Post irrigation soil and vegetation samples (if appropriate) were taken for chlorpyrifos residue analysis



53 and Time Domain Reflectrometry was used to measure water infiltration rates into the soil. The intensity of the synthetic storm was approximately 1.1 in/hr for two hours (i.e., an event having a l-in-5 year return frequency for this section of Iowa). In addition, each plot was prewetted the day before an application was to be made to bring the surface soil moisture up to near field capacity. Lorsban* 15G insecticide T-band and Lorsban 4E insecticide broadcastincorporated experiments were performed at corn planting at maximum labeled rates. Lorsban 15G banded over the corn whorl and Lorsban 4E surface broadcast experiments (with corn present) were also performed approximately 40 days after planting. Application rates The watershed was treated with chlorpyrifos at typical use rates (1992) and maximum labeled rates (1993). Chlorpyrifos applications for the watershed are given in Table II. Application rates for the external meso-plots are given in Table III.

Residue Analysis

Field samples were analyzed by Dow AgroSciences at the Midland Michigan, or Indianapolis, Indiana locations. Analyses were performed using gas chromatography (GC) with electron capture detection (ECD), flame photometric detection (FPD), or Mass Spectroscopy (MS). Limit of detection (LOD) and limit of quantitation (LOQ) are given in Table IV.

Soil Sampling Methodology The surface soil was sampled and analyzed for chlorpyrifos at select time intervals throughout the study. The watershed was divided into six (1992) or four (1993) equal area subregions within the watershed (see Figure 6). Soil samples taken from each subregion were analyzed for chlorpyrifos residues. Variability in chlorpyrifos soil dissipation was characterized by keeping each subregion sample unique (i.e., no compositing).

* Trademark of Dow AgroSciences


54 Table II. Application rates and dates for chlorpyrifos applications made to the watershed Corn Stage


At Planting Mid-Late Whorl Canopy near closure At Planting Mid-Late Whorl

Method

Date

Formulation

Rate applied to watershed

T-Band

5/12/92 (Julian day 133)

Lorsban 15G

1.30

Banded over whorls Aerial Broadcast

T-Band Foliar/ Surface Broadcast

6/9/92 (Julian day 161) Lorsban 15G 7/14/92 (Julian day 196) Lorsban 15G 5/19/93 (Julian day 139) 6/29/93 (Julian day 180)

1.35 0.94

Lorsban 15G

2.12

Lorsban 4E

1.01

M

Following the T-band application, either a 4"x4"x38" (1992) or 4 x4"xl2" (1993) transect of surface soil (centered and placed perpendicular to the row) was taken for analytical determination of chlorpyrifos residues. The length of the transect was decreased in 1993 to avoid problems associated with working with such a large soil sample. The transect depth was 1" following all non Tband applications. Chlorpyrifos residue analysis of surface soil provides information about the mass of pesticide available prior to a precipitation/runoff event. Deep soil core samples to thirty-six inches from the soil surface were taken and analyzed for chlorpyrifos residues at specific sampling intervals even though the physicochemical properties of chlorpyrifos indicate the molecule was relatively immobile in soil. Residue analysis of subsurface soil samples provided mass balance closure for chlorpyrifos fate. A n acetate lined hydraulic sampler was used to obtain soil samples up to 36" in depth in a corner of the watershed from a randomized grid pattern. The first 0-4" of soil was removed by placing a metal sleeve directly into the soil and manually removing this soil layer. Once removed, the hydraulic probe was placed inside the metal sleeve and the remainder of the soil core to 36" (i.e., 4"-36") was taken.


55

Table 1H. Application rates for sprinkler irrigated meso-plots located outside the watershed boundaries (1992). Plot


1

Formulation Lorsban 15G

2

Lorsban 4E

3

Lorsban 4E

4

Lorsban 15G

Rate 2.07 lb A I acre 31b AI acre 1.5 lb AI acre 1.04 lb A I acre

Application Method T-Band

Planting Date May 16

Treatment Runoff Date Date May 17 May 16

Broadcast / Incorporate into soil Broadcast over corn

May 18

May 18

May 19

May 15

June 25

June 26

May 12

June 24

June 25

Band over top of corn

Table IV. Level of Detection and Level of Quantification for chlorpyrifos in multiple environmental matrices.

Matrix Runoff Water Runoff Sediment Pond Water Pond Sediment Bulk soil Transects Soil cores Corn

LOQ l.lOng/mL 0.040 jLig/g 62.9 pg/mL 1.18 ng/g 0.011 ng/g 0.012 |Lig/g 0.016 ng/g

WD 0.33 ng/mL 0.012 ^ig/g 18.9 pg/mL 0.35 ng/g 0.003 ng/g

o.oo4 ng/g 0.005 ng/g

CONCLUSIONS A comprehensive nested field study has been designed and implemented to address many of the limitations found in current U S E P A FIFRA Subdivision N guidelines. Hydrology, erosion, pesticide runoff, drift, degradation, and


56


leaching are measured. Information on crop growth, heterogeneous soil properties, and meteorological conditions are quantified. Meso-plot experiments were designed and implemented to address length scale issues when extrapolating results to the watershed scale. The study design includes attributes to address mass balance closure and scaling effects, all while being performed under Good Laboratory Practices. Results from this study provide a comprehensive data set useful for site-specific model validation and subsequent extrapolation to predict pesticide behavior for other diverse watershed systems. Field observation results are found elsewhere (1).

REFERENCES 1.

2.

3.

4.

5.

Cryer, S. A . , H . E . Dixon-White, C. K . Robb, P. N. Coody, J. White "An Integrated Approach for Quantifying Pesticide Dissipation Under Diverse Conditions: II) Field Study Observations". In Phelps, Winton, and Effland, ed. American Chemical Society Symposium Series, Bridging the Gap Between Environmental Fate Laboratory and Field Dissipation Studies in the Registration Process, American Chemical Society Books, 2001 Cryer, S. A. " A n Integrated Approach for Quantifying Pesticide Dissipation Under Diverse Conditions: III) Site Specific Model Validation Using G L E A M S , EPICWQ, and E X A M S " . In Phelps, Winton, and Effland, ed. American Chemical Society Symposium Series, Bridging the Gap Between Environmental Fate Laboratory and Field Dissipation Studies in the Registration Process, American Chemical Society Books, 2001 Cryer, S. A. and P. L . Havens. "An Integrated Approach for Quantifying Pesticide Dissipation Under Diverse Conditions: IV) Scaling and Regional Extrapolation". In Phelps, Winton, and Effland, ed. American Chemical Society Symposium Series, Bridging the Gap Between Environmental Fate Laboratory and Field Dissipation Studies in the Registration Process, American Chemical Society Books, 2001 Wauchope, R. D., R. L. Graney, S. Cryer, C. Eadsforth, A . W. Klein, and K . D. Racke. Pesticide Runoff: Methods and Interpretation of Field Studies. Pure & Appl. Chem., 1995, Vol. 67, No. 12, pp. 2089-2108. Sharpley, A. N., and J. R. Williams, eds. EPIC-Erosion/Productivity Impact Calculator: 1. Model Documentation. U.S. Department of Agriculture Technical Bulletin. 1990, No. 1768. 235 pp.


Pesticide Environmental Fate - American Chemical Society

Recommend Documents